AI in fulfilment


According to economists at banking giant Goldman Sachs, ChatGPT and other artificial intelligence (AI) chatbots will put 300 million jobs at risk worldwide. 18% of all full-time jobs could end up being automated – and in particular the administrative tasks of white-collar professionals. Blue-collar workers in hands-on roles at distribution centres and elsewhere are less at risk of losing their jobs due to automation.

I recently visited FedEx’s newly enhanced international distribution centre (DC) in the Dutch town of Duiven. I was curious about the level of automation at this DC since the reopening. The first thing I noticed in the gigantic sorting hall was the sense of calm, the bright lighting and the relative quietness – which was especially remarkable in view of the fact that the central closed-loop sorting system handles 15,000 parcels per hour.

Manually stacking parcels

Despite the state-of-the-art automation, the DC operates three shifts totalling 410 people: 270 permanent staff and 140 temporary workers. The permanent employees (wearing yellow jackets) and the mostly Polish temporary workers seemed quite relaxed as they waited for the wave of parcels to arrive at the end of the afternoon. One important activity for them is loading the trucks: manually stacking parcels of all shapes and sizes to achieve the Key Performance Indicator of 50% capacity utilization.

It reminded me of a 3D version of the computer game Tetris that I played a lot as a student. Apparently, the AI has not yet reached the point where it can analyse the various weights and dimensions of the parcels to suggest the optimal stacking pattern to the warehouse operative. Instead, it’s still largely a complex puzzle that requires manual dexterity – and I don’t see robotic arms taking over this work in the near future.

Labour shortage

In contrast, the operations director at logistics service provider Active Ants recently told me that the company has opened a fifth fulfilment centre equipped with an automated storage system and autonomous mobile robots to process five million e-commerce orders per year. The mobile robots are guided through the fulfilment centre by AI-based software, and the operational efficiency will continue to improve thanks to machine learning. If extra capacity is required, additional robots can be deployed fairly easily. So although AI is penetrating the logistics domain at varying speeds, this is a real-life example of how it can offer solutions to the labour shortage.

Martijn Lofvers, Chief Trendwatcher Supply Chain Media